Predicting Count Loss in Modern Positron-Emission Tomography Systems

Abstract

The purpose of the data-acquisition electronics for any positron-emission tomography (PET) system is to detect and digitally encode annihilation events as they occur. Individual elements of the electronics are placed in parallel or cascade to organize the event information for subsequent processing. Each element is parameterized with a count loss L which is the fraction of events lost due to dead time (encoding delays, etc.) or data overflow in queueing circuits. This is an important parameter because the sensitivity of the tomograph in proportional to (1-L). The authors have categorized processing elements according to five device types. For each type, they find an expression for count loss. Some mathematical models that have appeared in the literature are applicable. These are extended here to include other devices, such as bank encoders and time-to-digital converters (TDC), with coincidence time resolving circuitry. Because some PET systems will have devices that do not fall into these categories, the authors show the derivations of the loss expressions so that one could easily extend their models with parallel derivations for other device types. In addition to PET systems, one should also be able to apply their results to other types of instruments which count random events.more » Although they concentrate mainly on count loss, they also briefly discuss the evaluation of other metrics of counting efficiency, which are the fraction of miscoded events and the fraction of ''random coincidence'' events.« less